Localization of semantic processing using functional magnetic resonance imaging

Publisher Website
Google Scholar
Abstract
The relationship between the functional components of language and the anatomic foci of their neural systems represents a central issue in cognitive neuroscience. Conflicting results from a number of laboratories using positron emission tomography (PET) imaging techniques have led to a significant controversy over the specific neuroanatomic sites engaged by semantic processing. We report here results of an experiment designed to address this controversy, that is, whether semantic processing activates temporal and/or frontal brain regions. In this experiment we used cognitive tasks that emphasized either semantic or phonological information processing but that were similar on both memory search and responce generation components, together with functional magnetic resonance imaging, to examine the neuroanatomic loci of lexical‐semantic as opposed to phonological processing. We studied nine right‐handed men performing two silent generation tasks: rhyme, and semantic category. The former focuses on word form (phonological information) while the latter focuses on word meaning (semantic information). By “phonological” we mean the process of apprehending the sound structures of language. By “semantic” we mean information about the word's contextually specified meanings. Semantic processing makes demands on, and activates widespread areas within, brain including the inferior frontal regions bilaterally and the left posterior temporal region. Phonological processing engages a more restricted neuroanatomic assembly involving primarily anterior left temporal lobe sites.